Estimation and Prediction in Dynamic Econometric Models
The presence of lags and other dynamic operators in the equation systems of economics is known to be a complicating factor in many aspects of estimation theory, but accepted practice is to follow the ‘comforting’ asymptotic result of Mann and Wald  that maximum likelihood estimates of linear difference equation systems are based on the specification that lagged variables may be formally treated like exogenous variables. Mann and Wald required dynamic stability, normality of error, stability of the distribution, and other assumptions that are usually made.
KeywordsExogenous Variable Prediction Horizon Prediction Efficiency Prediction Period Quadratic Loss Function
Unable to display preview. Download preview PDF.
- P. J. Dhrymes, A Simplified Structural Estimator for Large Scale Econometric Models, April, 1970, Discussion Paper No. 162, University of Pennsylvania.Google Scholar
- P. J. Dhrymes and V. Pandit, Asymptotic Properties of an Iterate of the Two Stage Least Squares Estimator, February, 1970, Discussion Paper No. 154, University of Pennsylvania.Google Scholar
- M. K. Evans and L. R. Klein, The Wharton Econometric Forecasting Model, 2nd ed. (Philadelphia: University of Pennsylvania, 1968).Google Scholar
- T. Haavelmo, The Probability Approach in Econometrics, Econometrica, XII (July 1944) supplement.Google Scholar
- A. L. Nagar, Statistical Estimation of Simultaneous Economic Relationships, Ph.D. Thesis, Rotterdam, 1959.Google Scholar
- H. Theil, Economic Forecasts and Policy, (Amsterdam: North-Holland, 1958).Google Scholar
- P. Whittle, Prediction and Regulation by Linear Least- Squares Method, (New York: D. Van Nostrand Company, Inc., 1963).Google Scholar